Exploring a Century of Meteorite Discoveries

Designing an interactive dashboard to explore 45,716 meteorite landings across 1,150+ years

R Programming

Shiny

Meteorite Landings

2025

From raw NASA records to interactive insight: a portfolio case study on building a professional R Shiny dashboard to explore global meteorite discoveries.

Author

Steven Ponce

Published

December 15, 2025

Explore the project

🚀 Live app: https://0l6jpd-steven-ponce.shinyapps.io/Meteorite_Landings/
💻 Source code: https://github.com/poncest/Meteorite_Landings

The Opportunity

NASA’s meteorite landings dataset documents 45,716 meteorite discoveries from 860 CE to 2013—over 1,150 years of recorded encounters with material from space.

This project began with a simple goal: turn a large historical dataset into an intuitive, interactive experience that allows users to explore where meteorites were found, how discoveries changed over time, and what distinguishes rare observed falls from later recoveries.

Guiding Questions

To shape the dashboard, I focused on a small set of questions that naturally emerge from the data:

Geography

Where are meteorite discoveries concentrated globally?
Why does Antarctica dominate modern records?

Time

How have discoveries evolved over time?
What explains the sharp rise in finds after 1970?

Physical Characteristics

How do observed falls differ from recovered meteorites?
Which meteorite types appear most frequently?

Design Approach: Progressive Disclosure

The dashboard follows a progressive disclosure strategy: present the big picture first, then allow deeper exploration for users who want it.

Global Overview

The opening view provides immediate context through four headline metrics:

Total meteorites: 45,716
Largest find: Hoba (~60 tonnes)
Time span: 860–2013
Observed falls: 2.4%

Figure 1: Global Overview tab showing worldwide meteorite locations, discovery trends over time, and mass distributions for observed falls versus recovered finds.

These are paired with three core visualizations:

Interactive world map showing all meteorite locations with clustering for performance
Discovery timeline highlighting the rapid increase in finds since the 19th century
Mass distribution comparing observed falls versus recovered meteorites

A simple year-range filter allows users to focus on specific historical periods without overwhelming the interface.

Deep Dive Analytics

For more detailed exploration, the analytics tab adds multidimensional filters:

Year range
Mass range (log scale)
Meteorite classification
Discovery status (Fell vs. Found)

All visualizations and summary statistics update reactively, allowing users to explore relationships across time, mass, and type.

Figure 2: Deep Dive Analytics tab with multidimensional filtering by year, mass, meteorite type, and discovery status, updating all charts and metrics in real time.

Data Explorer

The final tab exposes the underlying dataset through an interactive table, supporting:

Search and column-level filtering
Sorting
CSV export for external analysis

Figure 3: Data Explorer tab providing a searchable, sortable table of all meteorite records with CSV export for further analysis.

Key Patterns Revealed by the Data

Antarctica’s Outsized Role

Modern meteorite discoveries are dominated by Antarctica. The ice sheet acts as a natural conveyor belt, concentrating meteorites in accessible regions. Dark stones contrast sharply against the ice, and cold, dry conditions preserve specimens exceptionally well.

The Post-1970 Discovery Surge

The dramatic rise in discoveries after 1970 reflects improved search efforts, not an increase in meteorite impacts. Antarctic expeditions, desert surveys, and systematic recovery programs fundamentally changed how meteorites are found.

Fell vs. Found

Only 1,097 meteorites (2.4%) were observed falling. These events are scientifically valuable because their arrival time is precisely known. Recovered meteorites, which make up the remaining 97.6%, are often larger and discovered long after impact.

Design Decisions That Improved Clarity

Consistency Over Color

Early versions of the dashboard used multiple accent colors for value boxes, creating visual noise. Switching to uniform white cards with subtle borders improved readability and gave the dashboard a more professional, consulting-style appearance.

Standardized Layout

Both main tabs follow the same structure:

Key metrics
Optional filters
Detailed charts

This consistency reduces cognitive load and helps users quickly understand how to navigate the dashboard.

Unified Interactivity

All charts use ggiraph for hover-based interaction, ensuring a consistent interaction model across maps, timelines, and distributions.

Technical Overview

Architecture & Stack

The app uses a modular Shiny architecture:

UI: ui_main.R, ui_tabs.R
Server: server.R, server_charts.R
Global setup: global.R

Core technologies:

R Shiny + shinydashboard
ggiraph (interactive charts)
leaflet (mapping)
DT (data tables)
dplyr / ggplot2 (data manipulation and visualization)

Deployment Lessons

Expand for Deployment Notes

Dependency Management - Spatial packages (terra, raster) require explicit declaration in DESCRIPTION - Indirect dependencies must be captured in renv.lock with renv::snapshot()

Performance Optimization
- Binary data formats (.rds) load ~10x faster than text formats (.csv) - Essential for meeting shinyapps.io’s 60-second startup limit

Reproducibility - Lock all dependencies early to avoid deployment surprises - Test with rsconnect::appDependencies() before deploying

Closing Takeaway

Note

This project reinforced a simple but powerful lesson: effective dashboards are designed, not assembled. Clarity, consistency, and restraint mattered more than adding features — both in code and in communication.

Session Info

Expand for Session Info

R version 4.4.1 (2024-06-14 ucrt)
Platform: x86_64-w64-mingw32/x64
Running under: Windows 11 x64 (build 26100)

Matrix products: default


locale:
[1] LC_COLLATE=English_United States.utf8 
[2] LC_CTYPE=English_United States.utf8   
[3] LC_MONETARY=English_United States.utf8
[4] LC_NUMERIC=C                          
[5] LC_TIME=English_United States.utf8    

time zone: America/New_York
tzcode source: internal

attached base packages:
[1] stats     graphics  grDevices datasets  utils     methods   base     

loaded via a namespace (and not attached):
 [1] htmlwidgets_1.6.4 compiler_4.4.0    fastmap_1.2.0     cli_3.6.4        
 [5] htmltools_0.5.8.1 tools_4.4.0       rstudioapi_0.17.1 yaml_2.3.10      
 [9] rmarkdown_2.29    knitr_1.49        jsonlite_1.8.9    xfun_0.49        
[13] digest_0.6.37     rlang_1.1.6       renv_1.0.3        evaluate_1.0.1

GitHub Repository

Expand for GitHub Repo

Access the GitHub repository here

References

Expand for References

NASA Meteorite Landings Dataset: https://catalog.data.gov/dataset/meteorite-landings
Shiny Documentation: https://shiny.posit.co/
ggiraph: https://davidgohel.github.io/ggiraph/
Leaflet for R: https://rstudio.github.io/leaflet/

Citation

BibTeX citation:

@online{ponce2025,
  author = {Ponce, Steven},
  title = {Exploring a {Century} of {Meteorite} {Discoveries}},
  date = {2025-12-15},
  url = {https://stevenponce.netlify.app/projects/standalone_visualizations/meteorite_explorer.html},
  langid = {en}
}

For attribution, please cite this work as:

Ponce, Steven. 2025. “Exploring a Century of Meteorite Discoveries.” December 15, 2025. https://stevenponce.netlify.app/projects/standalone_visualizations/meteorite_explorer.html.

--- title: "Exploring a Century of Meteorite Discoveries" subtitle: "Designing an interactive dashboard to explore 45,716 meteorite landings across 1,150+ years" description: "From raw NASA records to interactive insight: a portfolio case study on building a professional R Shiny dashboard to explore global meteorite discoveries." date: "2025-12-15" author: - name: "Steven Ponce" url: "https://stevenponce.netlify.app" orcid: "0000-0003-4457-1633" citation: url: "https://stevenponce.netlify.app/projects/standalone_visualizations/meteorite_explorer.html" categories: ["R Programming", "Shiny", "Meteorite Landings", "2025"] tags: [ "r-shiny", "dashboard", "nasa-data", "meteorites", "interactive", "spatial-analysis" ] image: "thumbnails/sa_2025-12-15.png" format: html: toc: true toc-depth: 4 code-link: true code-fold: true code-tools: true code-summary: "Show code" self-contained: true theme: light: [flatly, assets/styling/custom_styles.scss] dark: [darkly, assets/styling/custom_styles_dark.scss] editor_options: chunk_output_type: inline execute: freeze: true cache: true error: false message: false warning: false eval: true --- ```{r setup} #| label: setup #| include: false knitr::opts_chunk$set( dev = "png", fig.width = 9, fig.height = 10, dpi = 320 ) ``` **Explore the project** - 🚀 **Live app**: <https://0l6jpd-steven-ponce.shinyapps.io/Meteorite_Landings/> - 💻 **Source code**: <https://github.com/poncest/Meteorite_Landings> ------------------------------------------------------------------------ ## The Opportunity NASA’s meteorite landings dataset documents **45,716 meteorite discoveries** from **860 CE to 2013**—over 1,150 years of recorded encounters with material from space. This project began with a simple goal: **turn a large historical dataset into an intuitive, interactive experience** that allows users to explore where meteorites were found, how discoveries changed over time, and what distinguishes rare observed falls from later recoveries. ------------------------------------------------------------------------ ## Guiding Questions To shape the dashboard, I focused on a small set of questions that naturally emerge from the data: **Geography** - Where are meteorite discoveries concentrated globally? - Why does Antarctica dominate modern records? **Time** - How have discoveries evolved over time? - What explains the sharp rise in finds after 1970? **Physical Characteristics** - How do observed falls differ from recovered meteorites? - Which meteorite types appear most frequently? ------------------------------------------------------------------------ ## Design Approach: Progressive Disclosure The dashboard follows a **progressive disclosure** strategy: present the big picture first, then allow deeper exploration for users who want it. ### Global Overview The opening view provides immediate context through four headline metrics: - **Total meteorites**: 45,716 - **Largest find**: Hoba (\~60 tonnes) - **Time span**: 860–2013 - **Observed falls**: 2.4% ![Global Overview tab showing worldwide meteorite locations, discovery trends over time, and mass distributions for observed falls versus recovered finds.](https://raw.githubusercontent.com/poncest/Meteorite_Landings/main/screenshots/global-overview.png){#fig-1} These are paired with three core visualizations: 1. **Interactive world map** showing all meteorite locations with clustering for performance 2. **Discovery timeline** highlighting the rapid increase in finds since the 19th century 3. **Mass distribution** comparing observed falls versus recovered meteorites A simple year-range filter allows users to focus on specific historical periods without overwhelming the interface. ### Deep Dive Analytics For more detailed exploration, the analytics tab adds multidimensional filters: - Year range - Mass range (log scale) - Meteorite classification - Discovery status (Fell vs. Found) All visualizations and summary statistics update reactively, allowing users to explore relationships across time, mass, and type. ![Deep Dive Analytics tab with multidimensional filtering by year, mass, meteorite type, and discovery status, updating all charts and metrics in real time.](https://raw.githubusercontent.com/poncest/Meteorite_Landings/main/screenshots/analytics.png){#fig-2} ### Data Explorer The final tab exposes the underlying dataset through an interactive table, supporting: - Search and column-level filtering - Sorting - CSV export for external analysis ![Data Explorer tab providing a searchable, sortable table of all meteorite records with CSV export for further analysis.](https://raw.githubusercontent.com/poncest/Meteorite_Landings/main/screenshots/data-explorer.png){#fig-3} ------------------------------------------------------------------------ ## Key Patterns Revealed by the Data ### Antarctica’s Outsized Role Modern meteorite discoveries are dominated by Antarctica. The ice sheet acts as a natural conveyor belt, concentrating meteorites in accessible regions. Dark stones contrast sharply against the ice, and cold, dry conditions preserve specimens exceptionally well. ### The Post-1970 Discovery Surge The dramatic rise in discoveries after 1970 reflects **improved search efforts**, not an increase in meteorite impacts. Antarctic expeditions, desert surveys, and systematic recovery programs fundamentally changed how meteorites are found. ### Fell vs. Found Only **1,097 meteorites (2.4%)** were observed falling. These events are scientifically valuable because their arrival time is precisely known. Recovered meteorites, which make up the remaining 97.6%, are often larger and discovered long after impact. ------------------------------------------------------------------------ ## Design Decisions That Improved Clarity ### Consistency Over Color Early versions of the dashboard used multiple accent colors for value boxes, creating visual noise. Switching to **uniform white cards** with subtle borders improved readability and gave the dashboard a more professional, consulting-style appearance. ### Standardized Layout Both main tabs follow the same structure: 1. Key metrics 2. Optional filters 3. Detailed charts This consistency reduces cognitive load and helps users quickly understand how to navigate the dashboard. ### Unified Interactivity All charts use **ggiraph** for hover-based interaction, ensuring a consistent interaction model across maps, timelines, and distributions. ------------------------------------------------------------------------ ## Technical Overview ::: {.callout-tip collapse="true"} ### Architecture & Stack The app uses a modular Shiny architecture: - **UI**: `ui_main.R`, `ui_tabs.R` - **Server**: `server.R`, `server_charts.R` - **Global setup**: `global.R` **Core technologies**: - R Shiny + shinydashboard - ggiraph (interactive charts) - leaflet (mapping) - DT (data tables) - dplyr / ggplot2 (data manipulation and visualization) ::: ------------------------------------------------------------------------ ## Deployment Lessons ::: {.callout-note collapse="true"} ##### Expand for Deployment Notes **Dependency Management** - Spatial packages (`terra`, `raster`) require explicit declaration in DESCRIPTION - Indirect dependencies must be captured in `renv.lock` with `renv::snapshot()` **Performance Optimization** - Binary data formats (`.rds`) load ~10x faster than text formats (`.csv`) - Essential for meeting shinyapps.io's 60-second startup limit **Reproducibility** - Lock all dependencies early to avoid deployment surprises - Test with `rsconnect::appDependencies()` before deploying ::: ------------------------------------------------------------------------ ## Closing Takeaway ::: {.callout-note} This project reinforced a simple but powerful lesson: effective dashboards are designed, not assembled. Clarity, consistency, and restraint mattered more than adding features — both in code and in communication. ::: ------------------------------------------------------------------------ ## Session Info ::: {.callout-tip collapse="true"} ##### Expand for Session Info ```{r, echo = FALSE} #| eval: true #| warning: false sessionInfo() ``` ::: ## GitHub Repository ::: {.callout-tip collapse="true"} ##### Expand for GitHub Repo [Access the GitHub repository here](https://github.com/poncest/Meteorite_Landings) ::: ------------------------------------------------------------------------ ## References ::: {.callout-tip collapse="true"} ##### Expand for References - NASA Meteorite Landings Dataset: <https://catalog.data.gov/dataset/meteorite-landings> - Shiny Documentation: <https://shiny.posit.co/> - ggiraph: <https://davidgohel.github.io/ggiraph/> - Leaflet for R: <https://rstudio.github.io/leaflet/> :::